Skip to content

Instantly share code, notes, and snippets.

@ftnext

ftnext/dataframe_max_pranctice_see_advise-20180224.ipynb

Last active February 24, 2018 03:51
Show Gist options

  • Select an option

    Learn more about clone URLs
  • Save ftnext/2ba44f0e43898f5ddeeeae2076a2cb01 to your computer and use it in GitHub Desktop.

Select an option

Learn more about clone URLs
Save ftnext/2ba44f0e43898f5ddeeeae2076a2cb01 to your computer and use it in GitHub Desktop.
Download ZIP
Display the source blob
Display the rendered blob
Raw
dataframe_max_pranctice_see_advise-20180224.ipynb
{
"cells": [
{
"cell_type": "code",
"execution_count": 1,
"metadata": {},
"outputs": [],
"source": [
"import pandas as pd"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## 使うデータ"
]
},
{
"cell_type": "code",
"execution_count": 2,
"metadata": {
"scrolled": false
},
"outputs": [
{
"data": {
"text/html": [
"<div>\n",
"<style scoped>\n",
" .dataframe tbody tr th:only-of-type {\n",
" vertical-align: middle;\n",
" }\n",
"\n",
" .dataframe tbody tr th {\n",
" vertical-align: top;\n",
" }\n",
"\n",
" .dataframe thead th {\n",
" text-align: right;\n",
" }\n",
"</style>\n",
"<table border=\"1\" class=\"dataframe\">\n",
" <thead>\n",
" <tr style=\"text-align: right;\">\n",
" <th></th>\n",
" <th>id</th>\n",
" <th>no</th>\n",
" <th>config_id</th>\n",
" <th>salesdate</th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>0</th>\n",
" <td>1</td>\n",
" <td>1</td>\n",
" <td>1</td>\n",
" <td>2011-11-12</td>\n",
" </tr>\n",
" <tr>\n",
" <th>1</th>\n",
" <td>2</td>\n",
" <td>3</td>\n",
" <td>3</td>\n",
" <td>2011-12-03</td>\n",
" </tr>\n",
" <tr>\n",
" <th>2</th>\n",
" <td>3</td>\n",
" <td>2</td>\n",
" <td>1</td>\n",
" <td>2012-02-10</td>\n",
" </tr>\n",
" <tr>\n",
" <th>3</th>\n",
" <td>4</td>\n",
" <td>1</td>\n",
" <td>2</td>\n",
" <td>2012-01-04</td>\n",
" </tr>\n",
" <tr>\n",
" <th>4</th>\n",
" <td>5</td>\n",
" <td>2</td>\n",
" <td>3</td>\n",
" <td>2012-06-03</td>\n",
" </tr>\n",
" <tr>\n",
" <th>5</th>\n",
" <td>6</td>\n",
" <td>1</td>\n",
" <td>3</td>\n",
" <td>2011-08-04</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"</div>"
],
"text/plain": [
" id no config_id salesdate\n",
"0 1 1 1 2011-11-12\n",
"1 2 3 3 2011-12-03\n",
"2 3 2 1 2012-02-10\n",
"3 4 1 2 2012-01-04\n",
"4 5 2 3 2012-06-03\n",
"5 6 1 3 2011-08-04"
]
},
"execution_count": 2,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"df3 = pd.read_csv('select_max_sample-3.csv', sep=',', header=0, index_col=None)\n",
"df3"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# ほしいもの\n",
"\n",
"config_idでグループ化し、各グループをsalesdateが最新の行で代表させたい"
]
},
{
"cell_type": "code",
"execution_count": 3,
"metadata": {
"scrolled": true
},
"outputs": [
{
"data": {
"text/html": [
"<div>\n",
"<style scoped>\n",
" .dataframe tbody tr th:only-of-type {\n",
" vertical-align: middle;\n",
" }\n",
"\n",
" .dataframe tbody tr th {\n",
" vertical-align: top;\n",
" }\n",
"\n",
" .dataframe thead th {\n",
" text-align: right;\n",
" }\n",
"</style>\n",
"<table border=\"1\" class=\"dataframe\">\n",
" <thead>\n",
" <tr style=\"text-align: right;\">\n",
" <th></th>\n",
" <th>id</th>\n",
" <th>no</th>\n",
" <th>config_id</th>\n",
" <th>salesdate</th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>0</th>\n",
" <td>3</td>\n",
" <td>2</td>\n",
" <td>1</td>\n",
" <td>2012-02-10</td>\n",
" </tr>\n",
" <tr>\n",
" <th>1</th>\n",
" <td>4</td>\n",
" <td>1</td>\n",
" <td>2</td>\n",
" <td>2012-01-04</td>\n",
" </tr>\n",
" <tr>\n",
" <th>2</th>\n",
" <td>5</td>\n",
" <td>2</td>\n",
" <td>3</td>\n",
" <td>2012-06-03</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"</div>"
],
"text/plain": [
" id no config_id salesdate\n",
"0 3 2 1 2012-02-10\n",
"1 4 1 2 2012-01-04\n",
"2 5 2 3 2012-06-03"
]
},
"execution_count": 3,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"ans_df3 = pd.read_csv('select_max_answer-3.csv', sep=',', header=0, index_col=None)\n",
"ans_df3"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"3通りのやり方を助言いただいた\n",
"1. datetimeに型変換してidxmax()を利用する\n",
"1. 型変換せずにグループごとの最大値のインデックスを取得する\n",
"1. グループごとに条件に合う行を見つける\n",
"\n",
"https://qiita.com/ftnext/items/48b7baa49217fdc37ae4#comment-8f7d0214dd2abdc18977"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## 助言1: datetimeに型変換してidxmax()を利用 を試す"
]
},
{
"cell_type": "code",
"execution_count": 4,
"metadata": {},
"outputs": [],
"source": [
"# TypeError: reduction operation 'argmax' not allowed for this dtype となる\n",
"# idx = df3.groupby('config_id')['salesdate'].idxmax()"
]
},
{
"cell_type": "code",
"execution_count": 5,
"metadata": {
"scrolled": true
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"<class 'pandas.core.frame.DataFrame'>\n",
"RangeIndex: 6 entries, 0 to 5\n",
"Data columns (total 4 columns):\n",
"id 6 non-null int64\n",
"no 6 non-null int64\n",
"config_id 6 non-null int64\n",
"salesdate 6 non-null object\n",
"dtypes: int64(3), object(1)\n",
"memory usage: 272.0+ bytes\n"
]
}
],
"source": [
"df3.info()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"salesdateはobject型。\n",
"\n",
"→idxmaxを使うためには型の変換が必要になるようだ"
]
},
{
"cell_type": "code",
"execution_count": 6,
"metadata": {
"scrolled": true
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"<class 'pandas.core.frame.DataFrame'>\n",
"RangeIndex: 6 entries, 0 to 5\n",
"Data columns (total 4 columns):\n",
"id 6 non-null int64\n",
"no 6 non-null int64\n",
"config_id 6 non-null int64\n",
"salesdate 6 non-null datetime64[ns]\n",
"dtypes: datetime64[ns](1), int64(3)\n",
"memory usage: 272.0 bytes\n"
]
}
],
"source": [
"df3['salesdate'] = pd.to_datetime(df3['salesdate'])\n",
"df3.info()"
]
},
{
"cell_type": "code",
"execution_count": 7,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"config_id\n",
"1 2\n",
"2 3\n",
"3 4\n",
"Name: salesdate, dtype: int64"
]
},
"execution_count": 7,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"idx = df3.groupby('config_id')['salesdate'].idxmax()\n",
"idx"
]
},
{
"cell_type": "code",
"execution_count": 8,
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"<div>\n",
"<style scoped>\n",
" .dataframe tbody tr th:only-of-type {\n",
" vertical-align: middle;\n",
" }\n",
"\n",
" .dataframe tbody tr th {\n",
" vertical-align: top;\n",
" }\n",
"\n",
" .dataframe thead th {\n",
" text-align: right;\n",
" }\n",
"</style>\n",
"<table border=\"1\" class=\"dataframe\">\n",
" <thead>\n",
" <tr style=\"text-align: right;\">\n",
" <th></th>\n",
" <th>id</th>\n",
" <th>no</th>\n",
" <th>config_id</th>\n",
" <th>salesdate</th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>2</th>\n",
" <td>3</td>\n",
" <td>2</td>\n",
" <td>1</td>\n",
" <td>2012-02-10</td>\n",
" </tr>\n",
" <tr>\n",
" <th>3</th>\n",
" <td>4</td>\n",
" <td>1</td>\n",
" <td>2</td>\n",
" <td>2012-01-04</td>\n",
" </tr>\n",
" <tr>\n",
" <th>4</th>\n",
" <td>5</td>\n",
" <td>2</td>\n",
" <td>3</td>\n",
" <td>2012-06-03</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"</div>"
],
"text/plain": [
" id no config_id salesdate\n",
"2 3 2 1 2012-02-10\n",
"3 4 1 2 2012-01-04\n",
"4 5 2 3 2012-06-03"
]
},
"execution_count": 8,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"df3.loc[idx]"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### 読み込み時に型変換する場合"
]
},
{
"cell_type": "code",
"execution_count": 9,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"<class 'pandas.core.frame.DataFrame'>\n",
"RangeIndex: 6 entries, 0 to 5\n",
"Data columns (total 4 columns):\n",
"id 6 non-null int64\n",
"no 6 non-null int64\n",
"config_id 6 non-null int64\n",
"salesdate 6 non-null datetime64[ns]\n",
"dtypes: datetime64[ns](1), int64(3)\n",
"memory usage: 272.0 bytes\n"
]
}
],
"source": [
"df3_2 = pd.read_csv('select_max_sample-3.csv', sep=',', header=0, index_col=None, parse_dates=['salesdate'])\n",
"df3_2.info()"
]
},
{
"cell_type": "code",
"execution_count": 10,
"metadata": {
"scrolled": true
},
"outputs": [
{
"data": {
"text/html": [
"<div>\n",
"<style scoped>\n",
" .dataframe tbody tr th:only-of-type {\n",
" vertical-align: middle;\n",
" }\n",
"\n",
" .dataframe tbody tr th {\n",
" vertical-align: top;\n",
" }\n",
"\n",
" .dataframe thead th {\n",
" text-align: right;\n",
" }\n",
"</style>\n",
"<table border=\"1\" class=\"dataframe\">\n",
" <thead>\n",
" <tr style=\"text-align: right;\">\n",
" <th></th>\n",
" <th>id</th>\n",
" <th>no</th>\n",
" <th>config_id</th>\n",
" <th>salesdate</th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>2</th>\n",
" <td>3</td>\n",
" <td>2</td>\n",
" <td>1</td>\n",
" <td>2012-02-10</td>\n",
" </tr>\n",
" <tr>\n",
" <th>3</th>\n",
" <td>4</td>\n",
" <td>1</td>\n",
" <td>2</td>\n",
" <td>2012-01-04</td>\n",
" </tr>\n",
" <tr>\n",
" <th>4</th>\n",
" <td>5</td>\n",
" <td>2</td>\n",
" <td>3</td>\n",
" <td>2012-06-03</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"</div>"
],
"text/plain": [
" id no config_id salesdate\n",
"2 3 2 1 2012-02-10\n",
"3 4 1 2 2012-01-04\n",
"4 5 2 3 2012-06-03"
]
},
"execution_count": 10,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"df3_2.loc[df3_2.groupby('config_id')['salesdate'].idxmax()]"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## 助言2: 型変換せずにグループごとの最大値のインデックスを取得する を試す"
]
},
{
"cell_type": "code",
"execution_count": 11,
"metadata": {
"scrolled": false
},
"outputs": [
{
"data": {
"text/html": [
"<div>\n",
"<style scoped>\n",
" .dataframe tbody tr th:only-of-type {\n",
" vertical-align: middle;\n",
" }\n",
"\n",
" .dataframe tbody tr th {\n",
" vertical-align: top;\n",
" }\n",
"\n",
" .dataframe thead th {\n",
" text-align: right;\n",
" }\n",
"</style>\n",
"<table border=\"1\" class=\"dataframe\">\n",
" <thead>\n",
" <tr style=\"text-align: right;\">\n",
" <th></th>\n",
" <th>id</th>\n",
" <th>no</th>\n",
" <th>config_id</th>\n",
" <th>salesdate</th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>2</th>\n",
" <td>3</td>\n",
" <td>2</td>\n",
" <td>1</td>\n",
" <td>2012-02-10</td>\n",
" </tr>\n",
" <tr>\n",
" <th>3</th>\n",
" <td>4</td>\n",
" <td>1</td>\n",
" <td>2</td>\n",
" <td>2012-01-04</td>\n",
" </tr>\n",
" <tr>\n",
" <th>4</th>\n",
" <td>5</td>\n",
" <td>2</td>\n",
" <td>3</td>\n",
" <td>2012-06-03</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"</div>"
],
"text/plain": [
" id no config_id salesdate\n",
"2 3 2 1 2012-02-10\n",
"3 4 1 2 2012-01-04\n",
"4 5 2 3 2012-06-03"
]
},
"execution_count": 11,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"df2 = pd.read_csv('select_max_sample-3.csv', sep=',', header=0, index_col=None)\n",
"idx = df2.groupby('config_id')['salesdate'].apply(lambda x: x.sort_values(ascending=False).index[0])\n",
"df2.loc[idx]"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"各部ごとに動きの確認。"
]
},
{
"cell_type": "code",
"execution_count": 12,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"(1, 0 2011-11-12\n",
"2 2012-02-10\n",
"Name: salesdate, dtype: object)\n",
"---\n",
"(2, 3 2012-01-04\n",
"Name: salesdate, dtype: object)\n",
"---\n",
"(3, 1 2011-12-03\n",
"4 2012-06-03\n",
"5 2011-08-04\n",
"Name: salesdate, dtype: object)\n",
"---\n"
]
}
],
"source": [
"# グループ化したあとのsalesdateの情報が取れている\n",
"for group in df2.groupby('config_id')['salesdate']:\n",
" print(group)\n",
" print('---')"
]
},
{
"cell_type": "code",
"execution_count": 13,
"metadata": {
"scrolled": false
},
"outputs": [
{
"data": {
"text/plain": [
"config_id \n",
"1 2 2012-02-10\n",
" 0 2011-11-12\n",
"2 3 2012-01-04\n",
"3 4 2012-06-03\n",
" 1 2011-12-03\n",
" 5 2011-08-04\n",
"Name: salesdate, dtype: object"
]
},
"execution_count": 13,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"df2.groupby('config_id')['salesdate'].apply(lambda x: x.sort_values(ascending=False))"
]
},
{
"cell_type": "code",
"execution_count": 14,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"config_id\n",
"1 Int64Index([2, 0], dtype='int64')\n",
"2 Int64Index([3], dtype='int64')\n",
"3 Int64Index([4, 1, 5], dtype='int64')\n",
"Name: salesdate, dtype: object"
]
},
"execution_count": 14,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"# グループごとにsalesdateの降順で並べ替えた後のindexが取れる\n",
"df2.groupby('config_id')['salesdate'].apply(lambda x: x.sort_values(ascending=False).index)"
]
},
{
"cell_type": "code",
"execution_count": 15,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"config_id\n",
"1 2\n",
"2 3\n",
"3 4\n",
"Name: salesdate, dtype: int64"
]
},
"execution_count": 15,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"# グループごとにsalesdateの降順で並べ替えた後のindexのうち先頭のものを取得\n",
"# （=グループごとにsalesdateが最大値の行のindexを取得した）\n",
"df2.groupby('config_id')['salesdate'].apply(lambda x: x.sort_values(ascending=False).index[0])"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## 助言3: グループごとに条件に合う行を見つける を試す"
]
},
{
"cell_type": "code",
"execution_count": 16,
"metadata": {},
"outputs": [],
"source": [
"df4 = pd.read_csv('select_max_sample-3.csv', sep=',', header=0, index_col=None)"
]
},
{
"cell_type": "code",
"execution_count": 17,
"metadata": {
"scrolled": true
},
"outputs": [
{
"data": {
"text/plain": [
"config_id\n",
"1 2012-02-10\n",
"2 2012-01-04\n",
"3 2012-06-03\n",
"dtype: object"
]
},
"execution_count": 17,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"# グループごとのsalesdateの最大値\n",
"df4.groupby('config_id').apply(lambda x: x['salesdate'].max())"
]
},
{
"cell_type": "code",
"execution_count": 18,
"metadata": {
"scrolled": true
},
"outputs": [
{
"data": {
"text/plain": [
"config_id \n",
"1 0 False\n",
" 2 True\n",
"2 3 True\n",
"3 1 False\n",
" 4 True\n",
" 5 False\n",
"Name: salesdate, dtype: bool"
]
},
"execution_count": 18,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"# 各行が所属するグループのsalesdateの最大値に一致するか\n",
"df4.groupby('config_id').apply(lambda x: x['salesdate']==x['salesdate'].max())"
]
},
{
"cell_type": "code",
"execution_count": 19,
"metadata": {
"scrolled": false
},
"outputs": [
{
"data": {
"text/html": [
"<div>\n",
"<style scoped>\n",
" .dataframe tbody tr th:only-of-type {\n",
" vertical-align: middle;\n",
" }\n",
"\n",
" .dataframe tbody tr th {\n",
" vertical-align: top;\n",
" }\n",
"\n",
" .dataframe thead th {\n",
" text-align: right;\n",
" }\n",
"</style>\n",
"<table border=\"1\" class=\"dataframe\">\n",
" <thead>\n",
" <tr style=\"text-align: right;\">\n",
" <th></th>\n",
" <th></th>\n",
" <th>id</th>\n",
" <th>no</th>\n",
" <th>config_id</th>\n",
" <th>salesdate</th>\n",
" </tr>\n",
" <tr>\n",
" <th>config_id</th>\n",
" <th></th>\n",
" <th></th>\n",
" <th></th>\n",
" <th></th>\n",
" <th></th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>1</th>\n",
" <th>2</th>\n",
" <td>3</td>\n",
" <td>2</td>\n",
" <td>1</td>\n",
" <td>2012-02-10</td>\n",
" </tr>\n",
" <tr>\n",
" <th>2</th>\n",
" <th>3</th>\n",
" <td>4</td>\n",
" <td>1</td>\n",
" <td>2</td>\n",
" <td>2012-01-04</td>\n",
" </tr>\n",
" <tr>\n",
" <th>3</th>\n",
" <th>4</th>\n",
" <td>5</td>\n",
" <td>2</td>\n",
" <td>3</td>\n",
" <td>2012-06-03</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"</div>"
],
"text/plain": [
" id no config_id salesdate\n",
"config_id \n",
"1 2 3 2 1 2012-02-10\n",
"2 3 4 1 2 2012-01-04\n",
"3 4 5 2 3 2012-06-03"
]
},
"execution_count": 19,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"df4.groupby('config_id').apply(lambda x: x[x['salesdate']==x['salesdate'].max()])"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Cookbookの `ln:84` の書き方も参考になる。\n",
"\n",
"https://pandas.pydata.org/pandas-docs/stable/cookbook.html#grouping"
]
},
{
"cell_type": "code",
"execution_count": 20,
"metadata": {
"scrolled": true
},
"outputs": [
{
"data": {
"text/plain": [
"config_id\n",
"1 3\n",
"2 4\n",
"3 5\n",
"dtype: int64"
]
},
"execution_count": 20,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"# salesdateがグループで最大の行のidを取り出す\n",
"df4_2 = pd.read_csv('select_max_sample-3.csv', sep=',', header=0, index_col=None, parse_dates=['salesdate'])\n",
"df4_2.groupby('config_id').apply(lambda subf: subf['id'][subf['salesdate'].idxmax()])"
]
},
{
"cell_type": "code",
"execution_count": 21,
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"<div>\n",
"<style scoped>\n",
" .dataframe tbody tr th:only-of-type {\n",
" vertical-align: middle;\n",
" }\n",
"\n",
" .dataframe tbody tr th {\n",
" vertical-align: top;\n",
" }\n",
"\n",
" .dataframe thead th {\n",
" text-align: right;\n",
" }\n",
"</style>\n",
"<table border=\"1\" class=\"dataframe\">\n",
" <thead>\n",
" <tr style=\"text-align: right;\">\n",
" <th></th>\n",
" <th>id</th>\n",
" <th>no</th>\n",
" <th>config_id</th>\n",
" <th>salesdate</th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>2</th>\n",
" <td>3</td>\n",
" <td>2</td>\n",
" <td>1</td>\n",
" <td>2012-02-10</td>\n",
" </tr>\n",
" <tr>\n",
" <th>3</th>\n",
" <td>4</td>\n",
" <td>1</td>\n",
" <td>2</td>\n",
" <td>2012-01-04</td>\n",
" </tr>\n",
" <tr>\n",
" <th>4</th>\n",
" <td>5</td>\n",
" <td>2</td>\n",
" <td>3</td>\n",
" <td>2012-06-03</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"</div>"
],
"text/plain": [
" id no config_id salesdate\n",
"2 3 2 1 2012-02-10\n",
"3 4 1 2 2012-01-04\n",
"4 5 2 3 2012-06-03"
]
},
"execution_count": 21,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"# idxmaxをapplyから適用できたが、\n",
"# インデックスを取得するだけなら ln [7] のように単にidxmaxを使った方がスッキリしているように思う\n",
"idx = df4_2.groupby('config_id').apply(lambda subf: subf['salesdate'].idxmax())\n",
"df4_2.loc[idx]"
]
}
],
"metadata": {
"kernelspec": {
"display_name": "Python 3",
"language": "python",
"name": "python3"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.5.1"
}
},
"nbformat": 4,
"nbformat_minor": 2
}
@ftnext
Copy link
Author

ftnext commented Feb 24, 2018

https://qiita.com/ftnext/items/48b7baa49217fdc37ae4#comment-8f7d0214dd2abdc18977
にていただいたコメントの動作確認をしたnotebookです。

@ftnext
Copy link
Author

ftnext commented Feb 24, 2018

Revision2での変更内容：
助言2についてsalesdateカラムをdatatime型に変換したdf3を使っていたため、
助言2の内容が確認できていなかったことに気づいた。
そこでsalesdateカラムがobject型のdf2を用意し、動作確認を行った。

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment